Search

Wall-cooling effect in hypersonic boundary layers can significantly alter the near-wall turbulence behavior, which is not accurately modeled by traditional RANS turbulence models. To address this shortcoming, this paper presents a turbulence modeling approach for hypersonic flows with cold-wall conditions using an iterative ensemble Kalman method. Specifically, a neural-network-based turbulence model is used to provide closure mapping from mean flow quantities to Reynolds stress as well as a variable turbulent Prandtl number. Sparse observation data of velocity and temperature are used to train the turbulence model. This approach is analyzed using direct numerical simulation database for boundary layer flows over a flat plate with a Mach number between 6 and 14 and wall-to-recovery temperature ratios ranging from 0.18 to 0.76. Two training cases are conducted: 1) a single training case with observation data from one flow case, 2) a joint training case where data from two flow cases are simultaneously used for training. Trained models are also tested for generalizability on the remaining flow cases in each of the training cases. The results are also analyzed for insights to inform the future work towards enhancing the generalizability of the learned turbulence model.

In this study, we present a universal nonlinear numerical scheme design method enabled by multi-agent reinforcement learning (MARL). Different from contemporary supervised-learning-based and reinforcement-learning-based approaches, no reference data and special numerical treatments are used in the MARL-based method developed here; instead, a first-principle-like approach using fundamental computational fluid dynamics (CFD) principles, including total variation diminishing (TVD) and $k$-exact reconstruction, is used to design nonlinear numerical schemes. The third-order finite volume scheme is employed as the workhorse to test the performance of the MARL-based nonlinear numerical scheme design method. Numerical results demonstrate that the new MARL-based method is able to strike a balance between accuracy and numerical dissipation in nonlinear numerical scheme design, and outperforms the third-order MUSCL (Monotonic Upstream-centered Scheme for Conservation Laws) with the van Albada limiter for shock capturing. Furthermore, we demonstrate for the first time that a numerical scheme trained from one-dimensional (1D) Burger's equation simulations can be directly used for numerical simulations of both 1D and 2D (two-dimensional constructions using the tensor product operation) Euler equations. The working environment of the MARL-based numerical scheme design concepts can incorporate, in general, all types of numerical schemes as simulation machines.

In recent years, using neural networks to speed up the solving of partial differential equations (PDEs) has gained significant traction in both academic and industrial settings. However, the use of neural networks as standalone surrogate models raises concerns about the reliability of solutions due to their dependence on data volume, quality, and training algorithms, especially in precision-critical scientific tasks. This study introduces a novel "super-fidelity" method, which uses neural networks for initial warm-starting in solving steady-state PDEs, ensuring both speed and accuracy. Drawing from super-resolution concepts in computer vision, our approach maps low-fidelity model solutions to high-fidelity targets using a vector-cloud neural network with equivariance (VCNN-e), maintaining all necessary invariance and equivariance properties for scalar and vector solutions. This method adapts well to different spatial resolutions. We tested this approach in two scientific computing scenarios: one with weak nonlinearity, using low Reynolds number flows around elliptical cylinders, and another with strong nonlinearity, using high Reynolds number flows over airfoils. In both cases, our neural operator-based initialization significantly accelerated convergence by at least two-fold, without sacrificing accuracy, compared to traditional methods. Its robustness is confirmed across various iterative algorithms with different linear equation solvers. The approach also demonstrated time savings in multiple simulations, even including model development time. Additionally, we propose an efficient training data generation strategy. Overall, our method offers an efficient way to accelerate steady-state PDE solutions using neural operators without loss of accuracy, especially relevant in precision-focused scientific applications.

Neural network-based turbulence modeling has gained significant success in improving turbulence predictions by incorporating high--fidelity data. However, the interpretability of the learned model is often not fully analyzed, which has been one of the main criticism of neural network-based turbulence modeling. Therefore, it is increasingly demanding to provide physical interpretation of the trained model, which is of significant interest for guiding the development of interpretable and unified turbulence models. The present work aims to interpret the predictive improvement of turbulence flows based on the behavior of the learned model, represented with tensor basis neural networks. The ensemble Kalman method is used for model learning from sparse observation data due to its ease of implementation and high training efficiency. Two cases, i.e., flow over the S809 airfoil and flow in a square duct, are used to demonstrate the physical interpretation of the ensemble-based turbulence modeling. For the flow over the S809 airfoil, our results show that the ensemble Kalman method learns an optimal linear eddy viscosity model, which improves the prediction of the aerodynamic lift by reducing the eddy viscosity in the upstream boundary layer and promoting the early onset of flow separation. For the square duct case, the method provides a nonlinear eddy viscosity model, which predicts well secondary flows by capturing the imbalance of the Reynolds normal stresses. The flexibility of the ensemble-based method is highlighted to capture characteristics of the flow separation and secondary flow by adjusting the nonlinearity of the turbulence model.

Learning turbulence models from observation data is of significant interest in discovering a unified model for a broad range of practical flow applications. Either the direct observation of Reynolds stress or the indirect observation of velocity has been used to improve the predictive capacity of turbulence models. In this work, we propose combining the direct and indirect sparse data to train neural network-based turbulence models. The backpropagation technique and the observation augmentation approach are used to train turbulence models with different observation data in a unified ensemble-based framework. These two types of observation data can explore synergy to constrain the model training in different observation spaces, which enables learning generalizable models from very sparse data. The present method is tested in secondary flows in a square duct and separated flows over periodic hills. Both cases demonstrate that combining direct and indirect observations is able to improve the generalizability of the learned model in similar flow configurations, compared to using only indirect data. The ensemble-based method can serve as a practical tool for model learning from different types of observations due to its non-intrusive and derivative-free nature., Comment: 42 pages, 16 figures

Multiphase flows through reservoir rocks are a universal and complex phenomenon. Relative permeability is one of the primary determinants in reservoir performance calculations. Accurate estimation of the relative permeability is crucial for reservoir management and future production. In this paper, we propose inferring relative permeability curves from sparse saturation data with an ensemble Kalman method. We represent these curves through a series of positive increments of relative permeability at specified saturation values, which guarantees monotonicity within, and boundedness between, 0 and 1. The proposed method is validated by the inference performances in two synthetic benchmarks designed by SPE and a field-scale model developed by Equinor that includes certain real-field features. The results indicate that the relative permeability curves can be accurately estimated within the saturation intervals having available observations and appropriately extrapolated to the remaining saturations by virtue of the embedded constraints. The predicted well responses are comparable to the ground truths, even though they are not included as the observation. The study demonstrates the feasibility of using ensemble Kalman method to infer relative permeability curves from saturation data, which can aid in the predictions of multiphase flow and reservoir production.

Most turbulence models used in Reynolds-averaged Navier-Stokes (RANS) simulations are partial differential equations (PDE) that describe the transport of turbulent quantities. Such quantities include turbulent kinetic energy for eddy viscosity models and the Reynolds stress tensor (or its anisotropy) in differential stress models. However, such models all have limitations in their robustness and accuracy. Inspired by the successes of machine learning in other scientific fields, researchers have developed data-driven turbulence models. Recently, a nonlocal vector-cloud neural network with embedded invariance was proposed, with its capability demonstrated in emulating passive tracer transport in laminar flows. Building upon this success, we use nonlocal neural network mapping to model the transport physics in the k-epsilon model and couple it to RANS solvers, leading to a PDE-free eddy-viscosity model. We demonstrate the robustness and stability of the RANS solver with a neural network-based turbulence model on flows over periodic hills of parameterized geometries. Our work serves as a proof of concept for using a vector-cloud neural network as an alternative to traditional turbulence models in coupled RANS simulations. The success of the coupling paves the way for neural network-based emulation of Reynolds stress transport models.

In this work, we propose using an ensemble Kalman method to learn a nonlinear eddy viscosity model, represented as a tensor basis neural network, from velocity data. Data-driven turbulence models have emerged as a promising alternative to traditional models for providing closure mapping from the mean velocities to Reynolds stresses. Most data-driven models in this category need full-field Reynolds stress data for training, which not only places stringent demand on the data generation but also makes the trained model ill-conditioned and lacks robustness. This difficulty can be alleviated by incorporating the Reynolds-averaged Navier-Stokes (RANS) solver in the training process. However, this would necessitate developing adjoint solvers of the RANS model, which requires extra effort in code development and maintenance. Given this difficulty, we present an ensemble Kalman method with an adaptive step size to train a neural network-based turbulence model by using indirect observation data. To our knowledge, this is the first such attempt in turbulence modelling. The ensemble method is first verified on the flow in a square duct, where it correctly learns the underlying turbulence models from velocity data. Then, the generalizability of the learned model is evaluated on a family of separated flows over periodic hills. It is demonstrated that the turbulence model learned in one flow can predict flows in similar configurations with varying slopes.

Developing robust constitutive models is a fundamental and longstanding problem for accelerating the simulation of complicated physics. Machine learning provides promising tools to construct constitutive models based on various calibration data. In this work, we propose a neural operator to develop nonlocal constitutive models for tensorial quantities through a vector-cloud neural network with equivariance (VCNN-e). The VCNN-e respects all the invariance properties desired by constitutive models, faithfully reflects the region of influence in physics, and is applicable to different spatial resolutions. By design, the model guarantees that the predicted tensor is invariant to the frame translation and ordering (permutation) of the neighboring points. Furthermore, it is equivariant to the frame rotation, i.e., the output tensor co-rotates with the coordinate frame. We evaluate the VCNN-e by using it to emulate the Reynolds stress transport model for turbulent flows, which directly computes the Reynolds stress tensor to close the Reynolds-averaged Navier--Stokes (RANS) equations. The evaluation is performed in two situations: (1) emulating the Reynolds stress model through synthetic data generated from the Reynolds stress transport equations with closure models, and (2) predicting the Reynolds stress by learning from data generated from direct numerical simulations. Such a priori evaluations of the proposed network pave the way for developing and calibrating robust and nonlocal, non-equilibrium closure models for the RANS equations.

Partial differential equations (PDEs) play a dominant role in the mathematical modeling of many complex dynamical processes. Solving these PDEs often requires prohibitively high computational costs, especially when multiple evaluations must be made for different parameters or conditions. After training, neural operators can provide PDEs solutions significantly faster than traditional PDE solvers. In this work, invariance properties and computational complexity of two neural operators are examined for transport PDE of a scalar quantity. Neural operator based on graph kernel network (GKN) operates on graph-structured data to incorporate nonlocal dependencies. Here we propose a modified formulation of GKN to achieve frame invariance. Vector cloud neural network (VCNN) is an alternate neural operator with embedded frame invariance which operates on point cloud data. GKN-based neural operator demonstrates slightly better predictive performance compared to VCNN. However, GKN requires an excessively high computational cost that increases quadratically with the increasing number of discretized objects as compared to a linear increase for VCNN.

Aerosol effects on micro-/macro-physical properties of marine stratocumulus clouds over the Western North Atlantic Ocean (WNAO) are investigated using in-situ measurements and large-eddy simulations (LES) for two cold air outbreak (CAO) cases (February 28 and March 1, 2020) during the Aerosol Cloud meTeorology Interactions oVer the western ATlantic Experiment (ACTIVATE). The LES is able to reproduce the vertical profiles of liquid water content (LWC), effective radius r_eff and the cloud droplet number concentration Nc from fast cloud droplet probe (FCDP) in-situ measurements for both cases. Furthermore, we show that aerosols affect cloud properties (Nc, r_eff, and LWC) via the prescribed bulk hygroscopicity of aerosols and aerosol size distributions characteristics. Nc, r_eff, and liquid water path (LWP) are positively correlated to the bulk hygroscopicity of aerosols and aerosol number concentration (Na) while cloud fractional cover (CFC) is insensitive to the bulk hygroscopicity of aerosols and aerosol size distributions for the two cases. The changes to aerosol size distribution (number concentration, width, and the geometrical diameter) allow us to disentangle aerosol effects on cloud properties from the meteorological effects. We also use the LES results to evaluate cloud properties from two reanalysis products, ERA5 and MERRA-2. Comparing to LES, the ERA5 reanalysis is able to capture the time evolution of LWP and total cloud coverage within the study domain during both CAO cases while MERRA-2 underestimates them.

Direct pore-scale simulations of fluid flow through porous media are computationally expensive to perform for realistic systems. Previous works have demonstrated using the geometry of the microstructure of porous media to predict the velocity fields therein based on neural networks. However, such trained neural networks do not perform well for unseen porous media with a large degree of heterogeneity. In this study we propose that incorporating a coarse velocity field in the input of neural networks is an effective way to improve the prediction performance. The coarse velocity field can be simulated with a low computational cost and provides global information to regularize the ill-posedness of the learning problem, which is usually caused by the use of local geometries due to the computational resource constraints. We show that incorporating the coarse-mesh velocity field significantly improves the prediction accuracy of the fine-mesh velocity field by comparison to the prediction that relies on geometric information alone, especially for the porous medium with a large interior vuggy pore space. We also show the flexibility of training the network in using coarse velocity fields with various resolutions. The results suggest that even using coarse velocity field with a very low resolution, the predictions are still enhanced and close to the ground truths. The feasibility of the method is further demonstrated by testing the trained network on real rocks. This study highlights the merits of incorporating a coarse-mesh velocity field into the input for neural networks, which provides global, physics-based information for the model, thereby improving the model's generalization capability.

Large-eddy simulation (LES) is able to capture key boundary-layer (BL) turbulence and cloud processes. Yet, large-scale forcing and surface turbulent fluxes of sensible and latent heat are often poorly prescribed for LES simulations. We derive these quantities from measurements and reanalysis obtained for two cold air outbreak (CAO) events during Phase I of the Aerosol Cloud meTeorology Interactions oVer the western ATlantic Experiment (ACTIVATE) in February-March 2020. We study the two contrasting CAO cases by performing LES and test the sensitivity of BL structure and clouds to large-scale forcings and turbulent heat fluxes. Profiles of atmospheric state and large-scale divergence and surface turbulent heat fluxes obtained from the reanalysis data ERA5 agree reasonably well with those derived from ACTIVATE field measurements for both cases at the sampling time and location. Therefore, we adopt the time evolving heat fluxes, wind and advective tendencies profiles from ERA5 reanalysis data to drive the LES. We find that large-scale thermodynamic advective tendencies and wind relaxations are important for the LES to capture the evolving observed BL meteorological states characterized by the hourly ERA5 reanalysis data and validated by the observations. We show that the divergence (or vertical velocity) is important in regulating the BL growth driven by surface heat fluxes in LES simulations. The evolution of liquid water path is largely affected by the evolution of surface heat fluxes. The liquid water path simulated in LES agrees reasonably well with the ACTIVATE measurements.This study paves the path to investigate aerosol-cloud-meteorology interactions using LES informed and evaluated by ACTIVATE field measurements., Comment: Accepted for publication in Journal of the Atmospheric Sciences

Training data-driven turbulence models with high fidelity Reynolds stress can be impractical and recently such models have been trained with velocity and pressure measurements. For gradient-based optimization, such as training deep learning models, this requires evaluating the sensitivities of the RANS equations. This paper explores the use of an ensemble approximation of the sensitivities of the RANS equations in training data-driven turbulence models with indirect observations. A deep neural network representing the turbulence model is trained using the network's gradients obtained by backpropagation and the ensemble approximation of the RANS sensitivities. Different ensemble approximations are explored and a method based on explicit projection onto the sample space is presented. As validation, the gradient approximations from the different methods are compared to that from the continuous adjoint equations. The ensemble approximation is then used to learn different turbulence models from velocity observations. In all cases, the learned model predicts improved velocities. However, it was observed that once the sensitivity of the velocity to the underlying model becomes small, the approximate nature of the ensemble gradient hinders further optimization of the underlying model. The benefits and limitations of the ensemble gradient approximation are discussed, in particular as compared to the adjoint equations.

The emerging push of the differentiable programming paradigm in scientific computing is conducive to training deep learning turbulence models using indirect observations. This paper demonstrates the viability of this approach and presents an end-to-end differentiable framework for training deep neural networks to learn eddy viscosity models from indirect observations derived from the velocity and pressure fields. The framework consists of a Reynolds-averaged Navier-Stokes (RANS) solver and a neural-network-represented turbulence model, each accompanied by its derivative computations. For computing the sensitivities of the indirect observations to the Reynolds stress field, we use the continuous adjoint equations for the RANS equations, while the gradient of the neural network is obtained via its built-in automatic differentiation capability. We demonstrate the ability of this approach to learn the true underlying turbulence closure when one exists by training models using synthetic velocity data from linear and nonlinear closures. We also train a linear eddy viscosity model using synthetic velocity measurements from direct numerical simulations of the Navier-Stokes equations for which no true underlying linear closure exists. The trained deep-neural-network turbulence model showed predictive capability on similar flows.